The broad goal of this proposal is to apply recent developments in the theory and measurement of emotion to investigation of crucial links between alcohol consumption and affective response and to explore variations in this relationship across individuals. Central to this effort is the conceptualization of emotion as a disposition (readiness) to approach or avoid, described in terms of two orthogonal dimensions of response: valence (directionality of response) and arousal (intensity of response). Measurement of the effect of drinking on subjective and behavioral components of these dimensions is part of the comprehensive assessment called for in this proposal. However, the most innovative feature of the research strategy is inclusion of psychophysiological indicators of emotional disposition that tap not only the effect of alcohol on nonspecific autonomic nervous system arousal and reactivity, but also measure affective valence directly. This would be accomplished by employing the "startle-probe paradigm," a variation of methods used extensively in the animal literature to evaluate fear and the actions of pharmacological agents on its expression. In human subjects, the magnitude of a reflex eyeblink (startle response) elicited by an unexpected noise burst (probe) is reliably augmented when introduced in the context if unpleasant stimuli, but attenuated during the processing of pleasant stimuli, thereby providing an unbiased index of emotional valence. This experiment would capitalize on the startle modulation effect to examine the supposed anxiolytic and mood elevating properties of alcohol, both thought to play a key role in motives for drinking. Variations in alcohol and startle modulation as a function of individual differences associated with vulnerability to alcoholism would also be explored. The overall approach is compatible with a biopsychosocial view of drinking and results of the work could have important applications in the prevention and treatment of alcohol problems.